SummaryThe author has studied 478 sera with the indirect immunofluorescence method and the classical serological methods (Bordet-Wasserman, Reiter, Kahn, V.D.R.L.) and Nelson’s test (TIT). The results of all these agreed for 442 sera (92,4 %), i.e. for 109 posittive and 334 negative sera. But in 36 cases (7,6 %) the results of the immunofluorescence disagree with those of the classical serology or TIT : the IF results were positive 24 times, negative 12 times. But in 33 cases, the IF raction agreed with the clinical diagnosis or with the TIT.In three sera only, the IF reactions seemed to be wrong. They were sera no. 7 (IF negative), no. 27 (IF positive ++) and no. 23 (IF positive + [+]). Taking into account the TIT and eventually the clinical symptoms, the results of the IF test are correct in 475 cases out of 478, and may be false in the 3 other Cases (i.e. 0,3 % of the total).Among the tests of the classical serology, when the IF test is positive, it is the V.D.R.L. test which is most often also positive.In the routine analysis, the IF test can economize material, personnel and time. But its chief advantage is surely a very great certainty, because it is a very sensitive and specific test and it closely agrees with the TIT and the clinical picture.